| { | |
| "algo": "APPO", | |
| "env": "quadrotor_multi", | |
| "experiment": "00_baseline_see_0", | |
| "experiments_root": "quad_single_baseline/baseline_", | |
| "help": false, | |
| "train_dir": "./train_dir", | |
| "device": "gpu", | |
| "seed": 0, | |
| "save_every_sec": 120, | |
| "keep_checkpoints": 3, | |
| "save_milestones_sec": 10000, | |
| "stats_avg": 100, | |
| "learning_rate": 0.0001, | |
| "lr_schedule": "constant", | |
| "lr_schedule_kl_threshold": 0.008, | |
| "train_for_env_steps": 10000000000, | |
| "train_for_seconds": 10000000000, | |
| "obs_subtract_mean": 0.0, | |
| "obs_scale": 1.0, | |
| "gamma": 0.99, | |
| "reward_scale": 1.0, | |
| "reward_clip": 10.0, | |
| "encoder_type": "mlp", | |
| "encoder_subtype": "mlp_quads", | |
| "encoder_custom": "quad_multi_encoder", | |
| "encoder_extra_fc_layers": 0, | |
| "hidden_size": 16, | |
| "nonlinearity": "tanh", | |
| "policy_initialization": "xavier_uniform", | |
| "policy_init_gain": 1.0, | |
| "actor_critic_share_weights": false, | |
| "use_spectral_norm": false, | |
| "adaptive_stddev": false, | |
| "initial_stddev": 1.0, | |
| "experiment_summaries_interval": 20, | |
| "adam_eps": 1e-06, | |
| "adam_beta1": 0.9, | |
| "adam_beta2": 0.999, | |
| "gae_lambda": 1.0, | |
| "rollout": 128, | |
| "num_workers": 36, | |
| "recurrence": 1, | |
| "use_rnn": false, | |
| "rnn_type": "gru", | |
| "rnn_num_layers": 1, | |
| "ppo_clip_ratio": 0.1, | |
| "ppo_clip_value": 5.0, | |
| "batch_size": 1024, | |
| "num_batches_per_iteration": 1, | |
| "ppo_epochs": 1, | |
| "num_minibatches_to_accumulate": -1, | |
| "max_grad_norm": 5.0, | |
| "exploration_loss_coeff": 0.0, | |
| "value_loss_coeff": 0.5, | |
| "kl_loss_coeff": 0.0, | |
| "exploration_loss": "entropy", | |
| "max_entropy_coeff": 0.0, | |
| "num_envs_per_worker": 4, | |
| "worker_num_splits": 2, | |
| "num_policies": 1, | |
| "policy_workers_per_policy": 1, | |
| "max_policy_lag": 100000000, | |
| "traj_buffers_excess_ratio": 1.3, | |
| "decorrelate_experience_max_seconds": 10, | |
| "decorrelate_envs_on_one_worker": true, | |
| "with_vtrace": false, | |
| "vtrace_rho": 1.0, | |
| "vtrace_c": 1.0, | |
| "set_workers_cpu_affinity": true, | |
| "force_envs_single_thread": true, | |
| "reset_timeout_seconds": 120, | |
| "default_niceness": 0, | |
| "train_in_background_thread": true, | |
| "learner_main_loop_num_cores": 1, | |
| "actor_worker_gpus": [], | |
| "with_pbt": false, | |
| "pbt_mix_policies_in_one_env": true, | |
| "pbt_period_env_steps": 5000000, | |
| "pbt_start_mutation": 20000000, | |
| "pbt_replace_fraction": 0.3, | |
| "pbt_mutation_rate": 0.15, | |
| "pbt_replace_reward_gap": 0.1, | |
| "pbt_replace_reward_gap_absolute": 1e-06, | |
| "pbt_optimize_batch_size": false, | |
| "pbt_optimize_gamma": false, | |
| "pbt_target_objective": "true_reward", | |
| "pbt_perturb_min": 1.05, | |
| "pbt_perturb_max": 1.5, | |
| "use_cpc": false, | |
| "cpc_forward_steps": 8, | |
| "cpc_time_subsample": 6, | |
| "cpc_forward_subsample": 2, | |
| "with_wandb": true, | |
| "wandb_user": null, | |
| "wandb_project": "sample_factory", | |
| "wandb_group": null, | |
| "wandb_job_type": "SF", | |
| "wandb_tags": [ | |
| "swarm-rl", | |
| "sf1", | |
| "single" | |
| ], | |
| "benchmark": false, | |
| "sampler_only": false, | |
| "env_frameskip": 1, | |
| "env_framestack": 4, | |
| "pixel_format": "CHW", | |
| "quads_discretize_actions": -1, | |
| "quads_clip_input": false, | |
| "quads_effort_reward": null, | |
| "quads_episode_duration": 15.0, | |
| "quads_num_agents": 1, | |
| "quads_neighbor_hidden_size": 256, | |
| "quads_neighbor_encoder_type": "no_encoder", | |
| "quads_collision_reward": 5.0, | |
| "quads_collision_obstacle_reward": 0.0, | |
| "quads_settle_reward": 0.0, | |
| "quads_settle": false, | |
| "quads_vel_reward_out_range": 0.8, | |
| "quads_settle_range_meters": 1.0, | |
| "quads_collision_hitbox_radius": 2.0, | |
| "quads_collision_falloff_radius": 4.0, | |
| "quads_collision_smooth_max_penalty": 10.0, | |
| "neighbor_obs_type": "none", | |
| "quads_use_numba": true, | |
| "quads_obstacle_mode": "no_obstacles", | |
| "quads_obstacle_num": 0, | |
| "quads_obstacle_type": "sphere", | |
| "quads_obstacle_size": 0.0, | |
| "quads_obstacle_traj": "gravity", | |
| "quads_local_obs": -1, | |
| "quads_local_coeff": 1.0, | |
| "quads_local_metric": "dist", | |
| "quads_view_mode": "local", | |
| "quads_adaptive_env": false, | |
| "quads_mode": "mix", | |
| "quads_formation": "circle_horizontal", | |
| "quads_formation_size": 0.0, | |
| "room_dims": [ | |
| 10, | |
| 10, | |
| 10 | |
| ], | |
| "quads_obs_repr": "xyz_vxyz_R_omega", | |
| "replay_buffer_sample_prob": 0.0, | |
| "anneal_collision_steps": 0.0, | |
| "quads_obstacle_obs_mode": "relative", | |
| "quads_obstacle_hidden_size": 32, | |
| "quads_collision_obst_smooth_max_penalty": 10.0, | |
| "quads_obst_penalty_fall_off": 10.0, | |
| "command_line": "--env=quadrotor_multi --train_for_env_steps=1000000000 --algo=APPO --use_rnn=False --num_workers=36 --num_envs_per_worker=4 --learning_rate=0.0001 --ppo_clip_value=5.0 --recurrence=1 --nonlinearity=tanh --actor_critic_share_weights=False --policy_initialization=xavier_uniform --adaptive_stddev=False --with_vtrace=False --max_policy_lag=100000000 --hidden_size=256 --gae_lambda=1.00 --max_grad_norm=5.0 --exploration_loss_coeff=0.0 --rollout=128 --batch_size=1024 --quads_use_numba=True --quads_mode=mix --quads_episode_duration=15.0 --quads_formation_size=0.0 --encoder_custom=quad_multi_encoder --with_pbt=False --quads_collision_reward=5.0 --quads_neighbor_hidden_size=256 --neighbor_obs_type=pos_vel --quads_settle_reward=0.0 --quads_collision_hitbox_radius=2.0 --quads_collision_falloff_radius=4.0 --quads_local_obs=6 --quads_local_metric=dist --quads_local_coeff=1.0 --quads_num_agents=8 --quads_collision_reward=5.0 --quads_collision_smooth_max_penalty=10.0 --quads_neighbor_encoder_type=attention --replay_buffer_sample_prob=0.75 --anneal_collision_steps=300000000 --train_for_env_steps=10000000000 --hidden_size=16 --neighbor_obs_type=none --quads_local_obs=-1 --quads_num_agents=1 --replay_buffer_sample_prob=0.0 --anneal_collision_steps=0 --save_milestones_sec=10000 --quads_neighbor_encoder_type=no_encoder --with_wandb=True --wandb_tags swarm-rl sf1 single --seed=0 --experiment=00_baseline_see_0 --train_dir=./train_dir --experiments_root=quad_single_baseline/baseline_", | |
| "cli_args": { | |
| "algo": "APPO", | |
| "env": "quadrotor_multi", | |
| "experiment": "00_baseline_see_0", | |
| "experiments_root": "quad_single_baseline/baseline_", | |
| "train_dir": "./train_dir", | |
| "seed": 0, | |
| "save_milestones_sec": 10000, | |
| "learning_rate": 0.0001, | |
| "train_for_env_steps": 10000000000, | |
| "encoder_custom": "quad_multi_encoder", | |
| "hidden_size": 16, | |
| "nonlinearity": "tanh", | |
| "policy_initialization": "xavier_uniform", | |
| "actor_critic_share_weights": false, | |
| "adaptive_stddev": false, | |
| "gae_lambda": 1.0, | |
| "rollout": 128, | |
| "num_workers": 36, | |
| "recurrence": 1, | |
| "use_rnn": false, | |
| "ppo_clip_value": 5.0, | |
| "batch_size": 1024, | |
| "max_grad_norm": 5.0, | |
| "exploration_loss_coeff": 0.0, | |
| "num_envs_per_worker": 4, | |
| "max_policy_lag": 100000000, | |
| "with_vtrace": false, | |
| "with_pbt": false, | |
| "with_wandb": true, | |
| "wandb_tags": [ | |
| "swarm-rl", | |
| "sf1", | |
| "single" | |
| ], | |
| "quads_episode_duration": 15.0, | |
| "quads_num_agents": 1, | |
| "quads_neighbor_hidden_size": 256, | |
| "quads_neighbor_encoder_type": "no_encoder", | |
| "quads_collision_reward": 5.0, | |
| "quads_settle_reward": 0.0, | |
| "quads_collision_hitbox_radius": 2.0, | |
| "quads_collision_falloff_radius": 4.0, | |
| "quads_collision_smooth_max_penalty": 10.0, | |
| "neighbor_obs_type": "none", | |
| "quads_use_numba": true, | |
| "quads_local_obs": -1, | |
| "quads_local_coeff": 1.0, | |
| "quads_local_metric": "dist", | |
| "quads_mode": "mix", | |
| "quads_formation_size": 0.0, | |
| "replay_buffer_sample_prob": 0.0, | |
| "anneal_collision_steps": 0.0 | |
| }, | |
| "git_hash": "7f0b112fb802e279ad83b0391d962b665dc44ba1", | |
| "git_repo_name": "https://github.com/Zhehui-Huang/quad-swarm-rl.git", | |
| "wandb_unique_id": "00_baseline_see_0_20220709_000003_519217" | |
| } |